Open Stack

Recently in Canberra (at OpenStack Day Australia - Gov't Edition), I
presented the Cloud Comparison Cost Model (C3M) created by Bruno+Bruno.

The response by the community was palpable as everyone in the audience was
taking pictures of the few short examples I borrowed from 2xBruno
presentation in Barcelona.

Naturally, there were several questions about how more dated could be
collected to populate the C3M with further examples.  I was not able to
answer this question, but would like to put it forward to the community
(esp the Bruno-ians) for how we might encourage additional data from both
public and private clouds?

Amy suggestions or modalities for building on this important work greatly
appreciated.

Kind Regards,

Flanders
---------- Forwarded message ----------
From: "David F Flanders" <flanders at openstack.org>
Date: 11 Nov. 2016 9:00 am
Subject: Re: Cost efffectiveness OpenStack
To: "Bruno Lago" <bruno at catalyst.net.nz>
Cc: "Bruno Morel" <bmorel at internap.com>, "Tom Fifield" <tom at openstack.org>

Excellent points both, I shall try and do my best to highlight this in the
keynote, and more importantly refer them to your excellent talk on the
subject.

Further to, we should be having this conversation on the user-committee
mailing list so as to get engagement by other providers who are willing to
plug their data into the cost model.  I would agree that this is going to
be difficult, but if framed the right way it can be presented so as to
highlight the transparency and openess which Internap and Catalyst are
upholding as exemplar open community members.
--> With your permission may I move this conversation over the UC mailing
list for presentation to the Large Deployment, Enterprise and Public Cloud
Working groups?

@Morel Re #cloudApp screencast, I've got some ideas based on the feedback
we had in the cloud app labs, more on this soon via the mailing list.

On Fri, Nov 11, 2016 at 8:12 AM, Bruno Lago <bruno at catalyst.net.nz> wrote:

> The only thing I'd like to point out is risk.
>
> Private cloud may be cheaper (at a certain scale), but can you attract and
> retain the right talent and pull out a complex project?
>
> I think people should take that in consideration as well. When you use a
> public cloud you pay more, but the risks are lower compared to having to
> build your own.
>
> Cheers,
> Bruno
>
> On Thu, 2016-11-10 at 16:31 +0000, Bruno Morel wrote:
>
> Hey David,
>
> I’m pretty sure Bruno’s gonna say it, but don’t forget that for the TCO of private cloud (what we called our ‘baseline’), Bruno used New Zealand hardware prices.
> We get a huge rebate in the US for example compared to all the shipping and handling necessary to move an Intel chip to beautiful Auckland :)
>
> We are completely in agreement on making that exercise a regular recurring occurrence, ideally with as much provider (public/private) as possible.
> I fear though than most of them won’t be so open. Openness is a culture that can be hard to realize inside some organizations (Rackspace going private may see more friction internally, for example).
>
> The link is still https://drive.google.com/drive/folders/0ByeoUtaR2My0b2g2YlJnTHhiNkU?usp=sharing
>
> Analyst are certainly looking into it, but they obviously want to be able to sell their own results on that subject (I know that the Forrester’s analyst I talked with is working on something like that).
> I would certainly not expect them to give it for free though :)...And I don’t know how much of our model they would be willing to reuse. For obvious economical reason, they will want to at least APPEAR to have done something on their own and ideally complicated so that people will want to pay a lot of money to get it :) We kinda took the opposite approach with Bruno, which is: simple, open and free...Don’t know how much they liked it to be honest :P
>
> In general, my opinion is that we will have more luck, as you said, with the different Working Groups (and I’d be interested to have the Large Deployment WG and the Enterprise WG look into the exercise too).
> Especially since visibility becomes much more easy to implement the moment the WG are included.
>
> So that’s where I’m ready to start the conversation with the other providers and try to make them do the exercice. At first probably internally and behind closed doors, and slowly convince them that, as long as they have their audience targeted correctly, there is a space for each of us...
>
> :)
>
> When are we doing our fre*king next screencast ? :P
>
> Bruno
>
> On 2016-11-09, 8:33 PM, "David F Flanders" <flanders at openstack.org> wrote:
>
>     Hola Brunos,
>
>     Tom and I are just preparing a keynote for the OpenStack (Gov't) Day
>     in Canberra.  We are hoping to highlight your excellent talk from the
>     Summit and was wondering if you have any other links, data or further
>     news from your discussions with the analysts?
>
>     Praises of yours which I am looking to sing, includes:
>
>      - Cost comparison done independently by two separate companies in NZ
>     & North America, comparing hardware resources (CPU, Storage, RAM) to
>     the equivalent data-centres in the same area.  Bandwidth is the most
>     expensive cost, and is highly dependent on local datacentre agreement
>     with the local broadband provider.
>      - The data for doing the comparison have been published openly for
>     reuse by anyone.  Other cloud companies have been invited to do
>     similar comparisons and publish their results, especially based on
>     different locations around the world.
>      - Baseline assumption are on a simple large web application with a
>     database, storage, etc.
>      - Comparisons have been done for both AWS vs Private Cloud OpenStack
>     (Catalyst) and AWS vs Public Cloud OpenStack (Internap).
>      - Practical real world example cost comparisons include: a.) large
>     cloud based website for marketing, b.) large cloud application, c.)
>     mobile cloud application, d.) disaster recovery scenario, e.) big data
>     workloads.
>      - Despite heavily weighting the comparison in favour of AWS pricing,
>     both OpenStack public and private clouds where more cost effective in
>     every category and example.
>      - Independent analysts are comparing the comparison data and it is
>     being validated.
>      - Cost savings can be provided by other OpenStack Public clouds in
>     comparison to AWS, especially depending your bandwidth usage per
>     region where your cloud application is being hosted. Significant cost
>     savings can be achieved by hosting your own cloud, but only at a
>     certain scale.
>
>     Greatly appreciated for any considerations or guidance.
>
>     On a personal note, I think promoting this dataset and getting other
>     public and private cloud providers to add their data is going to be a
>     key community activity moving forward, not least via the
>     PublicCloud-WG.  I hope to do my part, so please do let me know how I
>     can help you :)
>
>     Kind Regards,
>
>     Flanders
>     --
>     Flanders | OpenStack Foundation | Community Manager (Cloud Application
>     Communities)
>     http://superuser.openstack.org/articles/meet-openstack-s-community-wrangler-david-flanders
>
> --
> Bruno Lago
> Catalyst Cloudbruno at catalyst.net.nz+64 22 505-1234
>
>


-- 
Flanders | OpenStack Foundation | Community Manager (Cloud Application
Communities)
http://superuser.openstack.org/articles/meet-openstack-s-
community-wrangler-david-flanders
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/user-committee/attachments/20161115/0b4deda7/attachment-0001.html>